It’s so simple, a Data Scientist can do this.
- Alternatives to Advanced Data Analytics
- 4D Visualization
- Cycle Time
- Architectural Components
- Mobile App
- Client Programs Installation
- Open Excel Samples from Local Data Folder
- Udacity’s Business Analyst Nanodegree
- Data Transformation using Alteryx
- Date Adjustments in Excel
- Reshaper Add-in
- Save and Close Client Program
- Rename & Open Tableau Client Program
- File Types
- Load CSV from Public Sample Data Sets
- Add-ons to Tableau
- Public Datasets
- Public Galleries
- Choosing Attention
- Chart Types
- Getting Started Video
- Joins and Data Splits
- Distributing Content
- Google Analytics
- Tableau Server
- Tableau Extract API
- Master Tableau via User Certification
- Random links
- Show Tableau on Confluence
- More on front-end styling
This page provides an enhanced / deeper sequence to presenting Tableau based on a hands-on approach with less hype and marketing generalities.
- tableau.com/learn/starter-kit provides a learning sequence with clickable links to materials.
Tableau enables everyone to gain insights from raw data by providing a drag-and-drop UI working on data in many formats. In the biz it’s called a visual Data Analytics platform.
Alternatives to Advanced Data Analytics
Tableau’s promo video says:
- [0:30] “People build dashboards and applications like these, drag and drop, at up to 10 times the speed of a traditional business intellegence system”.
By “traditional” it means:
- IBM Cognos Insight,
- MicroStrategy Visual Insight,
- Oracle Exalytics appliance,
- SAS Visual Analytics Explorer,
- SAP Lumira ($995/user from SAP store)
- SAP Visual Intelligence
- Tibco Spotfire
Gartner notes “Tableau does not directly compete with other vendors with a strategy for being a broad BI platform offering. Tableau complements fully functional enterprise BI platforms that lack ADV capabilities.” That is why This Forrester report (although rather dated from 2012) has Tableau behind others on the strategy dimension:
The Gartner Magic Quadrant report in Feburary 2017 put Tableau almost neck-to-neck with Microsoft on the “ABILITY TO EXECUTE” scale:
In the previous year’s Magic Quadrant of Feb. 2016, Gartner’s kept Tableau at the top for Execution even though it downgraded most other vendors on that scale while it acknowledged Microsoft’s fast rise to be the leader in Completeness of Vision:
The Gartner Magic Quadrant report in 2015 also put Tableau above other vendors on their “ABILITY TO EXECUTE” scale.
Tableau focuses on presentation rather than both presentation and statistical calculations by software such as:
- R, an open-source (free) software
Since 2007 at http://www.gapminder.org/videos/ Hans Roling, a professor of health statistics in Sweden, is an internet legend for his “Joy of Stats” video shown on BBC Nov 10, 2010 and Ted Talks. In them he shows his Gapminder web app which presents multiple dimensions dynamically over time (nearly 300).
Here, different colors represent different countries, with the United States in red. Most other countries saw a decrease in fertility rate over time while life expectantcy increased.
This chart is an example of traits excellent visualization software possess:
Dynamic data - update visualizations (in “live mode”) as data changes in sources such as databases.
Visual querying - change the query by selecting or clicking on a portion of the graph or chart (to drill down, for example).
Linked multi-dimensional visualization - selections made in one chart are reflected as you navigate into other charts.
PROTIP: Dynamic (movie mode) is available only within the Tableau Public client, not when viewed on websites (as of 2016-01-05).
Personalization - give power users an in-depth view and newbies a simpler view, and also control access to data based on user- and role-based access privileges.
Actionable alerts - thresholds and parameters that trigger messages whether you’re interacting with reports or not.
Several editions of Tableau client programs can be installed on Windows and Mac laptops.
The licensed Desktop edition can read more types of data sources.
The Tableau Public website (http://public.tableausoftware.com) provides a gallery for showing workbooks created using the Public edition of client software. Tableau limits public workbooks up to 1 GB large containing up to 100,000 data rows.
Tableau also does not allow the Public client from storing files locally on a laptop.
PROTIP: Use the Public client when you have reliable public network access.
The Desktop edition can retrieve workbooks from the Public website. However, workbooks created on Tableau Desktop cannot be published to Tableau Public.
Private workbooks can be shared on a Tableau Server or in the Online service Tableau runs. A web app accessed using an internet browser is used to administer users on those servers.
Tableau’s native mobile apps on both Android and iPads make use of REST APIs that other servers can use to access data.
QUESTION: Can Mobile clients access Tableau Public?
https://gallery.alteryx.com/#! is viewed using a Alteryx Designer Windows client.
The Tableau Reader client can read (but not save) packaged workbooks.
Prices are subject to change at any time, and subject to negotiation. At time of writing, Tableau Desktop costs $1999. Tableau Server is around $1000 per user or ~$200K for an unlimited user license. Tableau Cloud costs $500/per user per year. The Tableau Cloud plan includes 100 GB of storage. I do not know the storage add-on costs for groups wanting more than 100 GB.
Tableau makes use of HTML5 within internet browsers. It’s unlike Microsoft Power Viewer, which depends upon installation of Silverlight and thus not mobile device friendly.
Tableau offers native mobile apps on both Android and iPads.
The apps reskins the internet browser, which can also be used to access workbooks.
- Click Sign In.
- For Sever Name, to use Tableau Public cloud, enter public.tableausoftware.com (without the https:// prefix). Ignore the “Tableau Online”.
- Click the orange Connect that appears to the right.
Client Programs Installation
The different Tableau clients are downloaded from separate URLs.
Tableau Desktop Personal and Pro editions are licensed at $999 and $1999 per year to access a wide range of data sources.
|Date||Component - Version||Installer||Installed|
|2017-05-20||TableauDesktop-10-2-2.dmg||463.8 MB||1.58 GB|
|2015-12-07||TableauReader-9-2-0.dmg||190.2 MB||668.4 MB|
|2015-12-01||TableauPublic-9-2-0.dmg||185.5 MB||683.8 MB|
|2015-07-20||TableauDesktop-9.2.0.dmg||192.1 MB||682.4 MB|
|2015-07-20||TableauDesktop-9.0.4.dmg||183.7 MB||643.4 MB|
QUESTION: Why is the Reader larger than others when it has less capability?
QUESTION: You don’t need to instal Reader if you have Deskstop installed?
The Server Download from here only runs within Windows (32 or 64 bit) and uses Active Directory Authentication.
There are several ways to open a Tableau client program:
- Open Excel Samples from Local Data Folder.
- Open a Tableau Client Program if you have your own file.
- Load CSV from Public Sample Data Sets.
Open Excel Samples from Local Data Folder
On a Mac, open a new Finder window.
Within your user’s Documents folder, scroll to click on the My Tableau Repository folder created during client installation.
Click to drill into the Datasources folder.
NOTE: The Tableau client remembers (encrypted) credentials to external data sources (such as Google Analytics) in this folder (as files with extension .tde).
Click on a version number of the Tableau client installed (9.2).
There is a separate folder for each version of Tableau (9.2, 9.0, etc.) because data structures often change with each new version.
Click on a locale (en_US_US).
Tableau clients 9.0 come with an Excel (.xls) file Sample - Superstore.xls. Notice it’s 3.4 MB.
Optionally, click on that file to open it within Excel (if you have Microsoft Office installed).
Within Excel, at the lower left corner, notice there are 3 tabs (worksheets) named Orders, Returns, People.
- Switch back to Finder. To open .xls files from the Finder, control-click on the xls file and select Open With > Other …, select Tableau.
- For Enable: select All Applications.
Select Tableau and click OK to view in Tableau client Data Connection window:
You can also get here by clicking “Excel” in Tableau’s Connect menu.
- Click the red x at the upper left corner to close the program.
Udacity’s Business Analyst Nanodegree
aka Predictive Analytics
Licenses are sent out AFTER the 7-day free period, when the first of $200 per month is paid.
Predicting Diamond Prices
In this initial project, students use results from a predictive model to make a recommendation on how much a jewelry company should bid for diamonds.
Multi-variate Linear Regression
To work through how a home-goods manufacturer can predict expected profits from a catalog launch, students build a linear regression model to provide results and a recommendation.
- Download the sample data .csv files.
In Finder, double-click on the diamonds.csv file to open it in Excel:
The training set contains historical data where the outcome is already known.
The sample diamonds.csv has 50,000 rows.
Outcome values are in the target variable (price).
Excel needs all variables together, so:
- Select the letter above the “caret” column to select the whole column.
- Press Ctrl+X to cut. Click on the column heading to the right of the “price” column.
Repeat for the “clarity_ord” and “cut_ord” columns.
- In Excel Data tab, Click on Data Analysis.
- If you don’t see “Data Analysis”, enable Analysis Toolpack. This only needs to be done once.
Select Regression and OK.
Conditions of attributes are in the predictor variables, also called dependent variables or (in Excel), Y Range.
- Click the red arrow icon for Y Range.
- Check “Labels”
Click in row 1 “price”. Press Command-shift-down to select the whole column
- Click the icon to collapse the pop-up.
- Click the icon to Output range.
- Cick cell M1.
Press Enter to accept it.
Variables with P-values (Prediction-values) beyond .15 should not be used.
Regression equation models enable predictions to be made based on attibute values.
The test dataset (new-diamonds.csv) contains new instances.
To create a multi-variate regression in Excel:
Select Location of a New Pet Store
Students use data preparation techniques to build a robust analytic dataset and use it to build a predictive model to select the best location for a pet store chain.
Data Transformation using Alteryx
NOTE: In 2015 Tableau did not have an ETL tool. Tableau add-in Alteryx for groups wanting true self-service ETL like Microsoft Power Query M scripting (formerly called Data Explorer) that does pivot/unpivot, joining, filtering, deduplicating, grouping, splitting, and transforms.
http://kb.tableausoftware.com/articles/knowledgebase/preparing-excel-files-analysis has some basic pivoting and cleansing features.
Alteryx performs Data blending of multiple and disparate data sources
Alteryx does predictive and spatial analytics with no coding required
Alteryx makes re-useable, modifiable workflows and direct updates to Tableau workbooks.
Date Adjustments in Excel
It may be easier to adjust data within Excel before Tableau uses the file, which is a number of seconds from 1970 (not the number since 1970 as Unix uses).
Tableau doesn’t read TIME according to (as of Oct 12, 2015) https://community.tableau.com/ideas/3315 and https://community.tableau.com/ideas/1054
Where A2 in the sample is replaced with wherever cell contains a date Excel recognizes, to convert the text “2016-01-12” to the internal date number of 42381.00:
Within Excel, hours and minutes are the decimals of a floating point number, such as .88 when the text equivalent is “21:40” on a 24 hour clock:
To convert 40 minutes into the hour, divide the number of minutes in a day (24 * 60 = 1440):
The number for minutes would not be more than 1/24 or 0.041666667.
To convert UTC/GMT times to local time that is -8 hours for Pacific Standard Time (Seattle):
NOTE UTC/GMT times are never adjusted for Daylight Savings, so the number of hours difference changes during the year. During the summer it’s 7 hours. See http://www.timeanddate.com/worldclock/converted.html?iso=20160114T00&p1=0&p2=234
http://www.cpearson.com/excel/TimeZoneAndDaylightTime.aspx The LocalOffsetFromGMT() function
To zero-fill leading zeroes:
File tableau_data_tool.zip downloaded from the website expands to Setup-Tableau-Add-In-7.0.exe. Thus it’s only for Windows edition of Office (not Mac).
Save and Close Client Program
After you click the red x at the upper left corner of the window, a pop-up appears if data was changed. A file name is suggested.
Change the name from Book2 or whatever is auto-assigned to something else.
If you change the default folder from Workbook (.twb) to Packaged workbook, data files and dashboard specifications will be packaged into a single .twbx file.
Load Local Sample Dashboards
With Tableau client UI program open:
Click menu Help Sample Workbooks. These were installed among binary files:
Superstore contains a forecast dashboard based on data connected live. This is different than what is presented in the Getting Started videos.
World Indicators (contains several dashboards referenced from a Story)
- Close the program.
Rename & Open Tableau Client Program
After installation on a Mac, open a Tableu client:
Open a Mac Finder window by holding down command and pressing Tab until the Finder icon is highlighted.
Pull down the Finder Go menu for Applications or press Shift+command+A.
The Tableau Desktop program is named simply “Tableau”.
NOTE: If a second version of the same program was installed, it would be automatically named “Tableau 2”.
PROTIP: Change the name to include version (such as 9.2.0 or 9.0.4, etc.).
Double-Click on the Tableau program to start it.
Return to the Start page by pressing command+2 or menu File.
On a Mac, click Tableau and About Tableau.
The top of the pop-up lists the specific version installed, such as “9.2.0 (9200.15.1201.0018) 64-bit””. Provide this to Tableau Support.
Click the red x at the upper left of the window to close the pop-up.
Notice the Tableau Desktop edition has “Server” in its menu:
EXTRA: More than one instance of the Tableau client can be running at the same time.
The above is from the Pro version, which has many drivers which connect to more data sources under the Connect section.
Click Statistical Files to accesses files from SAS and IBM SPSS:
NOTE: Not listed are MathCAD nor Microstrategy, SAP HANA, and other BI tools.
Select All files
As explained here:
.csv is a Comma Separated Value (character delimited) text format that Excel and many other programs open.
.xls is a Microsoft Office Excel 1998+ format.
.xlsx is a Microsoft Office Excel 2007+ format.
.txt files are also considerd among “Other files” that Tableau opens.
.tds data source files (as explained here) are like shortcuts containing information needed to just connect to the data sources, such as data source type, location, and custom fields.
.tdsx Packaged Data Source files contains all the information in the Data Source (.tds) file as well as any local file data sources (Excel, Access, text, and extracts). Packaged means a single zipped file for easier sharing a data source with people who may not have access to the original data that is stored locally on your computer.
.twb is the native workbook. It does not contain any data, which is why a packaged workbook is needed.
.twbx packaged workbook files (as explained here) contain one or more worksheets, plus zero or more dashboards and stories. These files are what is transferred to and from Tableau servers.
CAUTION: None of the above formats are encrypted with any data security.
.ttde are Shadow Extract files used for caching speed within clients and not directly referenced by users. Five of these files are stored within a Shadow Extract folder.
- PROTIP: You may want to create a separate folder to store all Excel data files locally in the same Excel folder. Store all Microsoft Office Access files in an Access folder.
Load CSV from Public Sample Data Sets
Use an internet browser (Firefox) to view where Tableau makes public datasets available.
- Scroll to The 2014 Inc. 5000.
Click on Dataset (csv) (comma separated values) link.
- Click the blue icon to the right of the Open With selection.
If Tableau (for the version you’re using) is not already listed, click Other and select it.
Click OK to save. The file would by default be saved to your user’s Download folder anyway.
CAUTION: All other openable files (.csv, .txt, .xls, etc.) are also shown.
[3:04] Click Sheet 1 at the bottom of the window to create a chart based on the data.
Close the program.
- PROTIP: In Finder, navigate to the Downloads folder and delete the .csv file. This clears off clutter.
See video http://www.tableau.com/learn/tutorials/on-demand/getting-started-data for more.
Add-ons to Tableau
Fine as it is out of the box, you should know about enhancements from others:
http://powertoolsfortableau.com/ from http://interworks.co.uk/business-intelligence/tableau-tools/
http://powertoolsfortableau.com/tools/workbook-tools-suite/ enables batch and automated interactions with Tableau workbooks (for $449 USD per year).
The $4999/year Workbook Software Development Kit (SDK) for Tableau by InterWorks enables batch, automated, programmatic access to Tableau workbooks. Visual Studio code completions. Semantic validations to catch breaking changes to your workbooks.
http://powertoolsfortableau.com/tools/portals-for-tableau enhances Tableau Server look-and-feel, filtering, and security.
Zion Spencer, InterWorks Software Sales Engineer 405/624-3214
Alteryx provides an alternative client.
The Client Pro certification exam uses the FAA Wildlife Strike Database
Tableau provides a Solution Guide which shows constructed viz to answer the certification questions:
* Do some species have more strikes in some months over others? * Are there any species appearing in the top that are surprising? * Calculate the strike rate (number of strikes divided by number of total flights) to see if the rate of wildlife strikes is increasing or decreasing over time. * Where are these strikes most likely to occur? * Are there certain months when wildlife strikes become more likely? * Have the number of strikes per year been changing over time? * Does the number of flights have any effect on seasonal trends? * Combine views into a dashboard allowing users to explore the data themselves. Note that you should not be confined to a specific number of species per month and should use a parameter to make a dynamic Top N.
Get the sample dataset:
- Scroll to the bottom of the page to
- Click <a target=”_blank” href=”http://wildlife.faa.gov/downloads/wildlife.zip</a> “FAA Wildlife Strike Database”</a> link without specifying any filters to download wildlife.zip.
- Unzip the file for a folder named “wildlife”.
- Drill into the folder.
- Double-click on “wildlife.accdb” to open it using Microsoft Office Access.
- The read-me.xls file in the folder is the same as the one obtained by clicking “total of 94 fields of data” for the fieldlist.xls. It contains a description of Column Names, Engine Codes, Aircraft Type, Engine Position.
If you are using a Mac, because Microsft Office for Mac 2011/2016 does not include Access, you’ll need to consider several options for reading the .accdb Microsoft Access database file.
The simplest option is to temporarily use a Windows machine or a virtual Windows machine containing MS Office Access – if they are already installed and available for use.
The $39.99 ODBC read-only driver for Microsoft Access from Actual Technologies enables Excel 2011 for Mac to execute SQL queries using Microsoft Query (comes with Microsoft Office) or a PivotTable.
- http://mthomas.co.uk/5-ways-to-access-access-on-a-mac/ notes that utilities such as http://mdbtocsv.codeplex.com support only older .MDB formats, not the .accdb format.
The $5.99 MDB Tool - For Microsoft Access by Hankinsoft Development Inc from Apple iTunes only handles databases in 2007 format, not 2010/2013 format.
To return data to a table: http://www.dummies.com/how-to/content/how-to-move-data-from-external-sources-into-excel-.html
To return PivotTable http://www.dummies.com/how-to/content/how-to-make-a-pivottable-in-excel-for-mac-2011.html
I think infographics are a form of art like paintings and sculptures. And beauty is in the eyes of the beholder. But clean looks are now preferred as being more modern.
PROTIP: To be good at creating your own, first get to know what others have done, starting from being able to name the different chart types.
Save the URL and click links to save workbooks to presentations that you might want to use.
WARNING: Most data in examples are several years out of date.
These galleries show why Tableau is among the most capable and popular of visualization software, providing a rich variety of chart types to make sense of data.
Tableau’s Gallery has URLs that are different with each new build of their website.
NOTICE: The permalink for the current version is near the bottom of each chart.
https://public.tableau.com/ is a social sharing site for people to show and blog about their visualizations of public data sets. This is much like what Adobe offers.
“VIZ of the day/week” from the Tableau Public sharing website is highlighted on the right side of opening landing page on Tableau clients.
QUESTION: How to describe changes over time? Videos instead of text blogs?
Vizzes by bloggers can be published on other sites, such as:
This “visual significance scale” is from SAP, but equally applicable for choosing how to convey a point:
The most attention is naturally given to the position of items, with items at the top given initial attention.
The value of various types are described in the video at https://www.youtube.com/watch?v=I_9nMSvY1FE which displays CO2 emissions by country, per-capita by decade, over time, by income group, grouped by region.
https://www.youtube.com/watch?v=amdR9SzcHDs&list=PLcTX_fKtDPs7x7DENN30sKjQvgwI1RtWW is a playlist of charts.
(new since 8.2) A sequence of points, each containing a short caption.
PROTIP: Have an initial point introducing the viz.
PROTIP: Pose a questions the content answers.
- https://public.tableau.com/profile/efields#!/vizhome/EarthquakesOnTheRise-Full/Earthquakestory enables users to select and provides a count of data ponints displayed on the map.
Boxes show relative sizes of components in the box, such as profit contribution, etc. (As long as data are all positive numbers)
provide a two-dimensional set of colors which add a third dimension.
show different size and colors in bubbles to show the relative size of components which don’t have anything in common with each other, such as sales from different industries.
show specific XY points around a trend line of central tendency:
PROTIP: Add the statistical r value such as “r=-0.74” to disclose extent of correlation.
Bar chart stacks
show the make-up of components:
show distribution of counts within a dataset.
are a form of bar line chart that
show the running total of each column
presents stages, as in productivity of various sales activities leading up to a sale.
Horizon (Trellis) charts
show when occassional “bumps” (such as errors) occur over time:
Used to show trends over time. Segments show different categories.
These have a thicker average line and upper and lower bounds.
Box and Whisker charts (new since v8) display quartiles in distributions:
Radar charts are great for visually identifying which of several dimensions are possessed by specific sub-populations shown using cicle-ish lines in the middle.
In this example, one can see that wines rated good and excellent (in dark red) have high alcohol while wines rated as poor (shown in thick black line) have low alcohol.
Geographic maps make use of geocodes that come with Tableau.
- Choropleth maps color land forms to differentiate values.
- Geo bubble charts overlay different size dots.
Tableau allows use of WMS servers, GIS sources and custom shapes
http://vizwiz.blogspot.com/p/dear-data-two.html shows how custom markers can be added on a map by editing the map’s CSV file.
In version 9 define custom territories off a map and Tableau merges on the fly. Streets and satellite maps from Mapbox.com are supported.
QUESTION: How about custom plots (such as fields on a farm).
- http://www.tableau.com/learn/tutorials/on-demand/custom-geocoding (7:03 minutes)
Getting Started Video
Tableau’s 23-minute Getting Started video shows step-by-step how to open a spreadsheet, create several charts from it, and a dashboard with a story.
But before diving in, get to know Keyboard Shortcuts
Download and Open the “Finished” Workbook
PROTIP: Start from the end. Download and click around the end product from the video before trying to follow the video step-by-step.
Control-click the Solution Workbook link to download the getting_started_finished.twbx workbook from
- In Finder, move the file to your Workbooks folder.
In Finder, click on the .twbx file to open within Tableau Desktop (not Tableau Public):
NOTE: The wrong workbook will be retrieved if you open the file using Tableau Public and click OK to message “A new version of this workbook is available. Tableau Public will open the latest version of this workbook.”
- On the Tableau menu, click Windows, Presentation Mode.
Click the Story view.
- Click story points from left
- Within “Overall, our profits look strong”, double-click on the red circle (Texas) to drill down.
- Click [-] to return to the world view.
- Click Category “Technology”.
- Click “But not across all categories”.
- Click “Here’s the biggest problem” for the bar chart.
- Cursor over the Tables red bar for more detail from Tooltip.
Click “What’s behind it?” dashboard.
Cursor over Customer Breakdown icons for details about each customer.
WARNING: Colors of the icons have not been specified as in the video.
- When you’re done, exit but don’t save.
Data: Global Superstore.xls
On a Mac, hold down fn key and click the Data Set link to download the global_superstore.zip file to your Downloads folder from http://www.tableau.com/sites/default/files/training/global_superstore.zip
WARNING: Do not use the xls file installed with Tableau 8 clients. Even though the file names are similar, it is missing the “Market” Dimension and “Shipping Cost” Measure referenced in the video.
Scroll to links to the Transcript pdf of the video.
- On a Mac, hold down fn key and click the Data Set link to download the global_superstore.zip file to your Downloads folder.
- In Finder, double-click on the file to unzip it to yield file Global Superstore.xls.
- Delete the global_superstore.zip file to unclutter your hard drive.
- Move the Global Superstore.xls file to your user local Documents / My Tableau Repository / Workbooks folder so that it appears after clicking Excel under Connect To a file.
Getting Started video content
PROTIP: Things happen fast on this video. So most beginners will need to pause frequently and watch several times. The web video presentation automatically jumps back to the beginning after long pauses, and doesn’t show the entire screen until it’s maximized. Annoying. So use link to save the MP4 video file so you can download it for viewing offline.
- Click on the video and register.
- Scroll to the MP4 link to the video that appears after registration.
Below are time codes to the video and links to my enhanced notes lower on this page.
- [0:30] Data Source (Global Superstore)
- [1:04] Connecting Live versus Extracting. Connect to the “Global Superstore.xls” file.
- [1:09] Joins and Data Splits
- [1:52] Rename field, Split Order ID Data Preparation split
- Dimensions and Measures
- Building Views
- Quick Table Calculations
[7:42] Exporting Data by right-click on the viz. Copy Data. Switch to Excel. Paste.
IMPORTANT: On a Mac, “right-click” by pressing down two fingers at a time on the touchpad. On Windows, it means clicking the right button on the mouse.
- [9:38] Global Sales Show Me chart types. Hold control key while clicking measures and dimensions.
- [11:08] Filters. Quick filter.
- [12:08] Hierarchies by dragging Sub-Category and dropping on Category. Call it Products.
[13:13] Expand and contract hierarchy clicking [+] and [-] on pills.
EXTRA: Remove a pill by dragging and dropping on the Dimensions list.
- [13:22] Swap axis and Fit Height
- [13:30] Sort from pill and axis.
- [14:05] Mark labels by clicking Abc button.
- [14:17] Drag Profits on top of Color Mark.
- [14:23] Group by dropping Market to the left of vertical axis.
- [14:31] Select headers to group similar items together using paperclip icon.
- [14:49] Edit Alias.
- [15:13] Customer Breakdown
- Working with Marks
- Trend Lines
- [17:50] Sales Dashboard
- [19:52] Story Points walk users through several captions on a navigator scroller, each analyzing a snapshop of visualizations.
- Callouts can be added on top of each visualization.
- [22:32] Distributing Content
Tableau’s VizQL is a visual query language which translates drag-and-drop actions into data queries expressed visually.
Joins and Data Splits
The sample Excel workbook contains Sheets named Orders, People, and Returns. “Users” are used in previous versions of this file.
Drag one of the Sheets (Orders) to the canvas pane.
Version 9+ automatically recognizes a “union (add rows)” drop target that concatenates multiple files containing the same headings.
Version 9+ has a Excel cleaning data interpreter feature that automatically recognizes and stiches together subtables into a single data source.
Drag another Sheet (Returns) to the canvas.
Tableau automatically suggests a join type:
If a column contains multiple fields, split them into sub-columns:
Because data is now loaded within Tableau, hover over a heading to Rename the heading to “Distribution Center”.
Delete split sub-columns not needed.
Click on the data type under the heading name to change it.
NOTE: You can later return to this by clicking tab “Data Source” at the lower left of the screen.
[3:06] Click on Sheet1 at the bottom set of tabs.
NOTE: Each sheet is also called a view.
You are asked to save the data.
EXTRA: [4:04] into video http://www.tableau.com/learn/tutorials/on-demand/getting-started-data shows linkage to add columns from another data source (Returns table) and inner vs left join.
Save Live or Extract Offline
[02:42] Specify a Live connection to dynamically retrieve fast-changing data or import a static Extract for offline usage.
CAUTION: Tableau does not save to disk automatically every few minutes (like Microsoft Office does).
If you save the file, the program suggests a name such as “Orders+ (Global Superstore)”.
PROTIP: Add a version number to the file name if you would like to be able to go back archival generations.
Shelves for Dimensions and Metrics
One of the amazing features of Tableau is it enables users to create sophisticated tables and charts just by dragging and dropping, then clicking around. (By contrast, some coding declarative statements used to define visualizations in Kibana.)
Each dimension element dragged to the Columns shelf defines a level in the hierarchy of columns.
[3:21] Drag from among Dimensions Category and drop within the Rows shelf.
The 3 category values are shown in a table.
“Pills” representing dimensions are color coded blue. [4:19] Dimensions are discreet categorical text strings displayed in labels.
[3:25] Drag from among Measures Quantity and drop within the Columns shelf.
“Pills” representing measures are color coded green. [4:09] Measures are usually continuous numbers such as counts, dollars, rankings, quantity, etc. These are the metrics we analyze and slice and dice.
[3:29] Drag from among Dimensions Segment and drop within the Rows shelf to the right of Category.
[3:32] Drag from among Dimensions Market and drop within the Colums shelf to the left of SUM(Quantity). Values are segregated across, starting with Africa.
[3:35] Drag from among Dimension Market again and drop it on top of Color (one of the Marks) makes use of Color to differentiate values.
EXTRA: Notice a Market legend appearing when.
EXTRA: Click on the Swap axis icon to swap X and Y axis (Columns and Rows).
[3:49] Notice Canada is an “emerging” market.
[4:41] The video shows a Clear Sheet by clicking the icon. But instead let’s control+click “Sheet 1” and select Rename the sheet to Market Sales.
Sales Seasonality View
EXTRA: Create a new worksheet.
[4:43] Drag Measure Sales and drop it on the left side. This is equivalent to dropping in the Rows shelf.
[4:49] Cursor over the blue bar to show a single result - the SUM of all sales.
[5:01] Drag Dimension Order Date and drop it on on the top box. This is equivalent to dropping in the Columns shelf.
[5:07] Click on the [+] in front of the Year (Order Date) pill to get Quarters, shown at the bottom of the view.
- [5:15] Drag Year to the right of Quarter so years are shown under each quarter.
- [5:26] Drag Year and drop on Color, which uses different colors instead of columns.
[5:28] Cursor to the right edge of the pill to select Months.
[5:46] Cursor to the right edge of the SUM(Sales) pill.
- Select Measure (Sum) to select Average (just to see what it looks like).
Press command+Z or click the Undo button.
[6:00] Cursor to the right edge of the SUM(Sales) pill and select Quick Table Calculations > Year over Year Growth.
NOTE: The pill now represents what was selected (Year over Year Growth).
- [6:23] Drag the SUM(Sales) pill and drop over Tooltip.
- [6:33] Drag Sales into Rows.
- [6:40] Hover over a line to see the tooltip.
- [6:53] Drage Categories to in front of SUM(Sales) in Rows.
[7:08] Annotate the yearly dip in sales during July by right-clicking a blank part of the viz and select Annotate | Area for the text box.
[7:25] Right-click on the viz and select Copy… Image, then click Copy.
[7:35] Right-click Sheet tab to Rename Sheet to Sales Seasonality.
QUESTION: Sesonality would be calculated statitcally how?
[7:41] Right-click on the viz to select Copy, Data, then press option+tab to be in Excel, and press command+V to paste into Excel.
EXTRA: Press command+S to save the worksheet.
[8:02] Right-click on sheet tab to select Duplicate as Crosstab.
This creates a new sheet.
- [8:04] Right-click on the Seasonality tab, select Duplicate as Crosstab.
- [8:11] Swap axis.
- Drag Category pill to the beginning of the Rows shelf.
- [8:14] Select Entire View from Normal.
- [8:15] Click Color button. Edit Colors
- [8:35] Drag Profit to drop on top of Color for clearer understanding of trends.
- Click Color Mark.
- Click Edit Colors.
- Check Use Full Color Range.
- Check Stepped Colors, 6 Steps. OK.
Undo the above tweaks.
- Change the Marks type to Square.
- Click Marks: Lables and choose Show mark labels (so text appears).
- Rename shee to Crosstab.
Global Sales and Profits
[9:08] Although Office Supplies is doing well, Furniture doesn’t have the same dark green swing. Where is it doing poorly? Is this happening across all stores across sales region?
- [9:18] Click icon or pull down menu to create new sheet.
- [9:38] We don’t know how to view the data. So make the Show Me menu visible.
- [10:05] Command-click multiple relevant Dimensions and Metrics for suggestions highlighted in the Show Me: Sales and Countries. Good matches for chart type are highlighted in blue.
- [10:22] Cursor among the highlighted chart icons.
[10:26] With simple map selected, click X to dismiss the Show Me menu.
This results in Columns: Longitude and Rows: Latitude.
- [10:27] Drag State and drop over map.
- [10:33] Click Size of Mark and increase size of dots on the map.
- [10:37] Click Color of Mark and change Transparency to 70%.
- Click on “Automatic” to the right of Borders within Marks and select a dark blue.
- [10:47] Click the gray bar of the Size legend to Hide Card.
[10:53] Drag Profit and drop on Color Marks to color the dots.
NOTE: The size of each circle is based on level of Sales.
But the color is based on level of profit. There are two facets presented.
- [10:56] Click the magnifier button inside the map for a geographic search.
- [11:00] Type United Kingdom in the search bar.
Click the unpin icon to zoom out.
- [11:13] Drag Categories to the Filters shelf for a pop-up.
- Select Furniture.
[11:25] Click on the right edge of the pill in Filters to select Show Quick Filter (Show Quick Filter in earlier versions) so users would have the option of choosing when viewing. See the choice appear on the right edge.
[11:29] Click the right edge of the Category filter to pull down and select
[11:46] command-click the right edge of a Dimension (Market) and select Show Filter to add another way users can select data interactively.
- [12:00] Double-click on the sheet name and Rename to Global Sales and Profit (but I prefer simply “Map”).
Sales by Category
[22:30] into Getting Started video, the display shown is hosted on localhost.
The 4 minute video http://www.tableau.com/learn/tutorials/on-demand/distributing-and-publishing provides a little more on this topic.
PROTIP: In the title and workbook name, specify a date reference since the work will be repeated in the future. Example: “Global Status 2015-01” because that’s the most recent data. Putting year before month enables sorting of text by name.
If you have been using the Tableau Public client, click on the Story before saving to the Tableau Public website.
Enter your password again. every. time.
A web page should pop up on your default internet browser.
- Click Edit in the green bar at the top.
Copy the URL from the browser, such as:
Obtain a permalink from http://goo.gl by pasting the URL.
PROTIP: There are many URL shortener utility sites. Google tracks link usage by all the Google Analytics metadata.
- Copy the generated permalink and paste it in the Permalink page.
Click the Share icon at the bottom.
Copy the Embed Code (example shown below) and embed it on your website:
An example of predictions is Bollinger Bands (described in video http://www.tableau.com/learn/tutorials/on-demand/bollinger-bands) based on calculation of moving averages going back periods of time and upper and lower bands based on various standard deviations.
[2:46] Alt (right) click on the white space within the Dimensions pane for its context menu. Select Create calculated Field. Define a new name and its formula.
[3:13] Edit to define a parameter to specify the periods to go back in time to calculate the Moving Average.
[8:26] Edit to define a parameter to specify the bands based on standard deviations.
TODO: Identify triggers to invoke when the moving average crosses one of the bands.
- https://www.youtube.com/watch?v=Y7OJjIeKcBM Tableau 8.1: Predictive Analytics with R
Data Blending is the combination of several data sets for integrated analysis. Although Tableau in 2015 lacked a data modeling diagram view, semantic model development is easy, rapid, intuitive and powerful with one or two combined data sources.
To specific options for a graph pane, click the caret at the right of the gray bar above that pane.
https://www.youtube.com/watch?v=bTU9sOfqjRo describes https://www.youtube.com/watch?v=CAZ3IAJEuCI Tips and Tricks from a Tableau Jedi by Alan Elderidge from Tableau Australia.
http://www.tableau.com/learn/tutorials/on-demand/connecting-google-analytics notes that extracts of Google Analytics are limited to 7 dimensions and 10 measures.
Tableau’s standard extraction from GA is limited in that it uses Sampler. So Google Big Query is a better approach for data extraction.
But Measure Groups pre-selects them
- Page Load Sample
- Page Load Time (ms)
- Time on Page
- Unique PageViews
The server enables workbooks to be organized into projects and assigned specific permissions by user group.
Sites isolate multiple tenants on the same hardware.
Tableau Server consists of several processes operating together at the same time (all Multi-threaded and Active/Active HA unless specified below):
|Application Server||wgserver.exe||log in, search, browse, generate static images, manage subscriptions|
|VizQL Server||vizqlserver.exe||Load and render requested views|
|Data Engine||tdeserver.exe, tdeserver64.exe||load Tableau Data Extracts in memory based on queries from VizQL processes|
|Backgrounder||backgrounder.exe||run maintenance tasks and data extract refreshes. Not multi-threaded as it tries to use as much as it can.|
|Data Server||dataserver.exe||handles requests to Tableau Data Sources. These requests can come from the Tableau Server or from Tableau Desktop users. 2x2.|
|Repository||postgres.exe||store settings, metadata, usage statistics and workbooks. Single process runs Active/Passive standby.|
|API Server||?||responds to API requests|
|Cache Server||?||holds items in memory|
|File Store||?||provides files|
Acess from views embedded in websites use Tableau’s “Trusted Tickets” authentication.
The chart above compares performance from v8.1 vs 8.0, obtained from:
See the V9 scalability report
Data for the chart above was obtained using Tableau’s TABCMD utility at
It’s not officially supported by Tableau, which is why many still use NeoLoad and LoadRunner, etc.
Summary of the tech:
- Works only on Windows
- Perfmon counters are enabled in a dataretrieve.config file
- Workbook viz to run are defind in CSV files (not in a GUI)
- Server test config are in .yml files
- JMX used to retrieve metrics
- Written using Java with SWT for the UI.
- Uses ProgresSQL to hold run results
- Custom automation needed to manage files before and after runs
Product Manager Neelesh Kamkolkar (@Neelesh Kamkol) has videos and blogs:
- https://www.youtube.com/watch?v=Ro9z3jSycIo – announcement TabJolt
- https://www.youtube.com/watch?v=Jzq9I-BrEfk – Getting Started
- https://www.youtube.com/watch?v=ty5a0RiSWDM – Load Testing
- https://www.youtube.com/watch?v=xrc8iLr1apw – Analyzing results
Planning and Architecting Tableau Server published Dec 4, 2014 notes use of SSO (Single Sign-On) support such as Microsoft Active Directory SAML refreshed using command-line utility tabcmd synchgroups.
- Tableau Server Administration and Authentication
Tableau Extract API
The 22 minute video at http://www.tableau.com/learn/tutorials/on-demand/extract-api-introduction explains the installation of Python 2.7.9 downloaded from http://www.tableau.com/data-extract-api which has both 32 and 64 bit binary Windows and Linux libraries using Python 2.7.9 or C/C++/Java.
[6:27] https://downloads.tableausoftware.com/tssoftware/TDE-API-Python-Linux-64Bit.gz expands to folder DataExtract-8300.15.0308.1149
[6:42] PROTIP: Create a Modules folder within Python27 to hold the DataExtract folder. cd to that folder and run the setup.py file:
python setup.py install
This creates files for this command:
>>> import dataextract >>> exit()
[8:47] Open IDLE installed with Python.
- http://vizninja.com/2015/11/15/how-to-properly-back-up-your-tableau-server/ by Paul Banoub
Tableau allows users to subscribe to dashboards for them to be automatically emailed to them on a schedule.
Master Tableau via User Certification
Exam Details & Prep Guides at http://www.tableau.com/support/certification provides information about taking Tableau’s tests Tests at conferences and on your laptop.
Tableau’s Qualified Associate for Tableau Desktop and Server ($250 to answer 16 questons over 2-hours).
Tableau’s Pro cert levels for Tableau Desktop and Server:
- http://mkt.tableau.com/files/Server-9-QA-Prep-Guide.pdf ($600 over 3 hours)
- http://mkt.tableau.com/files/Server-9-CP-Prep-Guide.pdf ($800 over 7 hours)
These pdf files contain links to just the front page (not specific links, unfortunately):
- (tutorial) http://www.tableau.com/learn/training
- (kb) = Knowledge-base : http://kb.tableau.com
- Product Help: http://onlinehelp.tableau.com
- Training from Tableau at http://www.tableau.com/support/training
More Trainings: http://www.tableau.com/learn/training?qt-training_tabs=1#qt-training_tabs
If you have a Pluralsight account: Ben Sullins (@bensullins, bensullins.com) created several video tutorials:
Business Dashboard Fundamentals Feb 19, 2014 - 3h 36m 42s provides advice about design of beautiful charts in Excel and Tableau 8.1, such as removing Chart Junk.
Data Analytics: Hands On Jul 15, 2015 - 5h 2m 19s provides an introduction to use of Excel and SQL and Tableau
Data Analysis Fundamentals with Tableau Sep 03, 2013 - 4h 47m 43s
Big Data Analytics with Tableau Jul 21, 2013 - 3h 44m 55s
Enterprise Business Intelligence with Tableau Server Nov 13, 2013 - 1h 36m 26s
Data Visualizations Using Tableau Public Sep 09, 2013 - 1h 47m 56s
This playlist of 15 videos from TechAnalysists begins with its Getting Started video.
https://www.youtube.com/watch?v=fEoB8EIEETQ describes Connecting to databases and advanced features.
Multidimensional data sources are supported by Tableau only in Windows, not in OSX.
Those what have taken the time to share their tips and tricks:
- http://www.dataplusscience.com/TableauReferenceGuide/ by Jeffrey Shaffer
- http://healthintelligence.drupalgardens.com/content/resources/tableau-training by Ramon Martinez
Tableau has hosted Iron Viz competitions where, over the course of a year, submissions are received for three online viz challenges, each with a different theme and judging criteria. The winner of each challenge earns a spot on stage at Tableau’s fall conference to compete in front of a live, data-savvy audience for the title of Iron Viz Champion.
http://data-dare.com/ is a fun way to evaluate visualizations.
#TC15 2015 (registration required)
\%TC16 in Austin, TX http://data-dare.com/2015/11/datadare1/#comment-116 run by Brit Cava who created https://public.tableau.com/profile/brit4337#!/vizhome/TC16AustinTXTravelGuide/AustinTravelGuide
https://itunes.apple.com/us/podcast/tableau-wanna-be-podcast/id908948125?mt=2 by Emily and Matt Francis who did TFF (Tableau French Festival)
Tableau was founded in 2003 by Christopher Stolte, Patrick Hanrahan, and Christian Chabot, now CEO.
VP Product Management Francois Ajenstat
VP Sales Kelly Wright will retire end of 2016.
Tableau corporate headquarters is in the bayside Freemont neighborhood of Seattle, Washington which locals call “The center of the universe”.
Tableau Software is traded on NYSE under ticker symbol DATA. The company went public on May 17, 2013 at an initial public offering price of $31 per share.
Tableau 10 for Essential Training 4h 22m by Curt Frye
Tableau 10 for Data Scientists 2h 4m by Matt Francis
This page is way too long and will be split up in separate pages … when I have time.
Show Tableau on Confluence
More on front-end styling
This is one of several topics:
- Text Editors
- Markdown text for GitHub from HTML
- JAM Stack Website Project Plan
- HTTP/2 Transition Project Plan
- Static websites
- Website styles
- Email from website
- Search within Hyde format Jekyll websites
- Data Visualization using Tableau